fal-shot-craft

Pass

Audited by Gen Agent Trust Hub on Nov 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: SKILL.md instructs the agent to "Check that the required local Python, Node, browser, renderer and media tools exist before using them," which implies executing shell commands to verify the local environment.
  • [EXTERNAL_DOWNLOADS]: The agent is instructed in SKILL.md to "Install publicly documented dependencies only within the user's authorized setup," creating a potential vector for downloading and installing third-party packages from external registries.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted user input in the form of video briefs or existing prompts to perform its primary function.
  • Ingestion points: User-provided briefs and existing prompts are ingested to be "improved" or converted into shot-craft descriptions (found in SKILL.md and production-guide.md).
  • Boundary markers: The instructions do not specify any delimiters or safety markers for processing external data.
  • Capability inventory: The skill allows for dependency installation, tool mapping (e.g., genmedia), and general command execution within the agent's host environment (SKILL.md).
  • Sanitization: There is no mention of sanitizing or validating user-provided prompts before processing or interpolating them into the workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 7, 2026, 07:17 AM